Machine Learning Engineer - Commerce Search

About the Team You'll Join The ML Engineer (Platform) at Toss Securities is part of the ML Platform Team within the Product Division. The mission of the ML Platform Team is to create an optimal machine learning platform that enables efficient and stable development and operation of various AI/ML services at Toss Securities. Your Responsibilities Upon Joining Develop and enhance the Gateway system, the gateway for ML services. Develop and operate a Gateway system based on FastAPI that handles enterprise-level LLM API requests. Design and implement authentication, routing, traffic control, fault isolation (Circuit Breaker, Fallback), large-scale TPS processing, and load balancing strategies from both application and infrastructure perspectives in the FastAPI-based Gateway application. Manage and serve ML services. Directly operate a machine learning model serving system in a Kubernetes environment. Design and improve the LLM serving architecture to operate stably under large traffic conditions. Monitor latency, error rates, resource usage, and analyze and resolve operational issues for the models in service. Identify root causes of failures and implement structural improvements, including operational policies and architecture. Develop and manage a common ML platform for the company. Develop and manage a common platform for efficiently operating the training and serving of internal ML/LLM models based on Kubeflow. Continuously monitor and optimize the performance and resources of workloads executed on the platform. Build infrastructure for LLM-based services. Operate LLM services using various serving frameworks such as vLLM, SGLang, and Triton. Manage the environment to ensure stable operation of training and serving workloads on high-performance GPU clusters like H100/B300. Build and operate a large-scale data training environment for finance domain-specific LLMs. We Are Looking for Candidates Who: Are proficient in one or more programming languages such as Python, Go, Java, or Kotlin, and have experience designing and developing API servers in production environments. Have experience developing or operating API Gateways (Nginx, Kong, etc.) or LLM Routers (LiteLLM, Envoy AI Gateway, etc.), with a background in handling high-volume traffic and incident response. Have experience operating serving logs and event pipelines integrated with Kafka, Elasticsearch, and Kibana. Have experience defining monitoring metrics for model serving and configuring and operating dashboards using Prometheus and Grafana. Have experience operating ML/LLM model serving using KServe, BentoML, vLLM, SGLang, etc. Have experience directly managing MLOps components (Kubeflow, KServe, Airflow, Argo CD, MLflow, etc.) in Kubernetes environments and debugging and resolving issues. Can design and apply long-term improvement plans through root cause analysis beyond short-term responses to issues that arise during service operations. Additional Preferred Experience: Experience in related fields or technologies will be a plus.

About the Team You'll Join The ML Platform Team Leader will oversee the pivotal ML Platform team responsible for the enterprise's AI/ML development and operational environment. This team comprises the 'ML Ops Part', which builds and operates a stable MLOps environment based on Kubernetes, and the 'LLM Part', which leads the development of cutting-edge LLM applications along with the necessary LLM Ops environment. Upon joining, you will oversee both Parts, establishing and executing a technical strategy that provides a scalable and efficient platform, allowing data scientists and ML engineers to focus solely on model development and experimentation. You will play a crucial role in presenting the vision for the ML Platform, fostering team growth, and solidifying the foundation of the company's AI technology capabilities. Your Responsibilities Establish and lead the technical vision and long-term roadmap for the ML Platform team (ML Ops Part, LLM Part). Mentor team members, conduct code reviews, and manage performance to drive team growth. Design and oversee the operation of a scalable and reliable ML Platform (model training, deployment, serving, monitoring) based on Kubernetes. Lead the technical development of various LLM applications and establish the LLM Ops environment for fine-tuning, serving, and evaluating large-scale models. Collaborate closely with cross-functional teams, including data scientists, ML engineers, and product teams, to resolve bottlenecks in AI/ML development and maximize developer experience (DX). Continuously research the latest MLOps and LLM technology trends, design the architecture of the enterprise AI tech stack, and provide technical direction. Define and continuously improve metrics for platform stability, cost efficiency, and performance optimization. Ideal Candidate We are looking for someone with proven experience in leading engineering teams (ML, Infrastructure, Platform, etc.), successfully presenting technical visions, establishing roadmaps, and nurturing team members. A deep practical experience in building and operating MLOps platforms (e.g., Kubeflow, MLflow, CI/CD, model serving) in a Kubernetes environment is essential. A strong understanding and practical experience in developing LLM applications (RAG, fine-tuning, agents, etc.) or LLM Ops (large-scale model serving, Vector DB, evaluation pipelines) is required. Experience in designing and optimizing system architecture capable of handling large-scale traffic and data is preferred. You should be able to define complex technical problems, communicate clearly with various stakeholders, and solve issues strategically. A passion for designing and advancing the platform while considering business impact and developer experience (DX) beyond mere technical adoption is essential. Resume Recommendations Detail your contributions to leading ML platform or system projects (technical leadership, architecture design, team management, strategic planning). Clearly outline the flow of addressing technical/organizational challenges for each project, including your proposed solutions (architecture, adopted technologies), collaboration processes with team members/stakeholders, and final results (platform performance improvements, acceleration in development speed, cost reductions). Instead of simply listing tasks, include insights and learnings gained as a technical leader, along with the criteria for making technical trade-offs. Emphasize quantifiable achievements that you have led. If you have experience successfully building/operating large-scale systems through collaboration with multiple engineering teams to resolve complex technical problems, please share that. Journey to Joining Toss Bank Application submission > Job interview > Cultural fit interview > Leadership interview > Reference check > Compensation negotiation > Final acceptance Please Note Any false information found in the resume or submitted documents may lead to disqualification.

Neural Concept creates AI-driven tools for engineering and design teams, with a focus on 3D deep learning and engineering intelligence. The company’s platform supports clients in automotive, aerospace, and energy as they improve product design processes. Based in Seoul, Neural Concept works with industry leaders to solve engineering challenges and speed up innovation. Role overview The Machine Learning Application Engineer role bridges engineering and artificial intelligence. This position is part of the ML Application Engineering team and involves direct collaboration with clients throughout Korea. The focus is on understanding each client’s technical needs and delivering practical, tailored solutions. Daily work includes close interaction with engineering and design teams as they adopt AI and machine learning in their product development cycles. What you will do Partner with customer engineering teams to manage projects, analyze and process engineering data using the Neural Concept platform and Python libraries, and develop custom workflows. Design and implement machine learning and deep learning solutions for engineering and physics challenges, including building proofs-of-concept for CAD, CAE, and manufacturing applications. Train client teams on Neural Concept’s platform and methods, ensuring smooth integration of AI into their workflows. Work with internal developers by sharing customer feedback and helping improve the product. Requirements Strong drive to address real-world engineering problems using machine learning and automation. Master’s or PhD in Engineering, Applied Mathematics, or Physics. Solid grasp of engineering fundamentals and machine learning techniques. Location Seoul, Korea

# Join Our Team- As a Machine Learning Engineer (MLE) for Home Recommendations, you will play a crucial role in optimizing the recommendation strategies for various content, services, promotions, and messages within the Toss app using machine learning.- Your work will involve precisely modeling the effectiveness of different content at various user touchpoints such as the top of the Toss home screen, banners after money transfers, and push notifications, while continuously improving actual recommendation outcomes.- To enhance the quality of recommendations across diverse areas, you will focus on quantitative performance forecasting and the design of sophisticated exposure strategies.# Responsibilities- Design and enhance recommendation models for content, services, and promotions displayed across the Toss app.- Develop models to predict appropriate recommendations and responses (e.g., CTR) based on user behavior, timing, and context.- Engage in iterative improvement of recommendation systems through feature discovery, experimental design, and performance evaluation.- Optimize exposure strategies for various recommendation areas and push messages based on user responses, quantitatively measuring actual performance.# Who We Are Looking For- Individuals with practical experience dealing with recommendation systems, ranking, and personalization modeling are essential.- A background in designing and experimenting with CTR prediction, user response prediction, and ranking modeling is crucial.- Experience in feature engineering and model performance enhancement using user behavior data is required.- Proficiency in major ML frameworks such as PyTorch, TensorFlow, and LightGBM is necessary.- Candidates who have improved actual service performance through iterative experimentation and quantitative analysis will be preferred.- The ability to clearly define problems and explain them technically while collaborating with diverse teams is highly valued.# Resume Tips- Please detail any projects you have worked on that have had significant organizational impact.- Specify the problems you defined, the approaches you chose, the experimental methods, and how you improved performance in modeling-centric projects.- Highlight any processes where you quantitatively solved problems through iterative experiments and performance analysis.# Application Process- Application Submission > 1st Technical Interview (Coding) > 2nd Technical Interview > Cultural Fit Interview > Reference Check > Compensation Negotiation > Final Acceptance- The first technical interview will include a simple coding test, resume check, and a basic ML knowledge assessment.- The second technical interview will involve in-depth technical questions and discussions on ML system design.# A Message for Future Colleagues- "This role is not just about modeling; it's about making an impact on the business."- The most satisfying aspect of working at Toss is that we do more than just modeling.- Previously, the work was limited to inputting data into existing models and evaluating performance, but now we focus on how to incorporate unaggregated data into our models.- It is rewarding to contribute to running a super app based on a deep understanding of users, moving beyond just analyzing and modeling financial-related data!

About UsJoin us in setting a global standard for AI in video understanding!At Twelve Labs, we are developing state-of-the-art AI models specialized in video processing, enabling advanced search, analysis, summarization, and insight generation from vast amounts of video data.Our models are utilized by the world's largest sports leagues to swiftly and accurately highlight content from extensive match footage, delivering an ultra-personalized viewing experience. Additionally, integrated control centers in South Korea leverage our technology for efficient CCTV video exploration to respond to crises swiftly. Major broadcasters and studios worldwide harness our models to create content for billions of viewers.Twelve Labs, a deep tech startup with offices in San Francisco and Seoul, has been recognized as one of the world's top 100 AI startups by CB Insights for four consecutive years. We have secured over $110 million in investments from leading VCs and corporations such as NVIDIA, NEA, Index Ventures, Databricks, and Snowflake. Our AI model, uniquely developed in Korea, is available through Amazon Bedrock. We thrive on collaboration with exceptional peers to create innovative products and grow alongside our global clientele.Our core values include:Honesty and reflection regarding ourselves and our team.A spirit of resilience and humility that doesn’t fear failure or feedback.A commitment to continuous learning to enhance team capabilities together.If you enjoy solving challenging problems and growing through collaboration, your opportunity awaits at Twelve Labs.About the TeamOur team focuses on multimodal representation learning and production serving. We integrate various modalities like video, audio, and text into a unified embedding space, ensuring stable serving to thousands of global customers.We conduct experiments on multimodal embedding models within a large-scale distributed learning environment, taking responsibility for the end-to-end process of transforming research findings into real-time inference systems. Leveraging top-tier GPU resources such as the NVIDIA B300, we minimize the transition cycle from research to production.In our rapid development cycles, we collaborate closely with Research, Product, and Infrastructure teams to create significant technical impacts, delivering research outcomes to customers globally within months.Role OverviewKey ResponsibilitiesDesign and optimize large-scale distributed learning pipelines for multimodal embedding models.Enhance the inference performance of embedding models in production environments (throughput, latency, cost-efficiency).Design and build vector search systems and embedding serving infrastructure.Improve and automate the ML pipeline covering model development, training, and serving to facilitate rapid transitions to production.Address applied research challenges such as data filtering and evaluation metric design to enhance model quality and user experience.Explore and experiment with AI-based development tools like Claude and Gemini to boost development productivity.Collaborate closely with Research, Product, and Infrastructure teams, taking ownership of the end-to-end process to deliver models to actual customers.Ideal Candidate ProfileResearch or development experience in computer vision, natural language processing, or multimodal learning.Proficiency in Python and PyTorch, with experience in model training within large-scale distributed environments.Experience with embedding models, vector search systems, and advanced ML techniques.

Welcome to the Journey of Joining the Daangn Team!At Daangn, we are committed to creating an environment where individuals can grow alongside the company's success.The Daangn recruitment team is here to ensure that you can engage in meaningful discussions with wonderful colleagues. Introducing the Software Engineer, Machine Learning RoleThe Software Engineer specializing in Machine Learning at Daangn leverages machine learning to enhance various services, ultimately improving user satisfaction and fostering connections within local communities. This role involves recommending personalized and engaging content for users on home feeds and detail pages, as well as facilitating efficient advertising to help local residents promote their offerings.Key Responsibilities[Feed Quality Team]Develop and enhance ML models to boost user engagement and foster diverse and serendipitous discoveries on the feed.Analyze data to formulate hypotheses, identify improvements based on theoretical foundations, and validate through online experiments to achieve tangible service impact.Develop and refine models for a deeper understanding of users and content.Utilize LLM in various ways to enhance personalized recommendations.[Advertising Recommendation Team]Develop and improve CTR and CVR prediction models, as well as automated bidding, targeting, and pacing models for ad recommendations and rankings.Analyze user behavior data and advertising performance, formulating data-driven hypotheses and validating them through A/B tests to enhance both advertising efficiency and user experience.Recommend meaningful ads to users through candidate ad extraction and ranking models that consider personalization.Enhance advertising targeting accuracy using LLM and the latest AI technologies, deepening the understanding of users and ad content.Who We're Looking ForIndividuals with a solid foundation in machine learning theory and principles.A deep understanding of deep learning (Recommendation, NLP, Graph Neural Networks, Reinforcement Learning, Vision, or more).Experience in conceptualizing efficient code architecture and writing readable code.Preferred QualificationsExperience making a tangible impact on user services through machine learning in a data-driven environment.A keen interest in the latest machine learning and deep learning trends, enjoying reading relevant papers.Experience with ML-based recommendation and advertising systems.Experience deploying machine learning models and systems in large-scale production environments.Familiarity with BigQuery, Cloud Dataflow, Kubeflow, TFX, and TF Serving.Additional InformationThis full-time position includes a 3-month probation period.In accordance with the 'Act on the Promotion of Employment for Persons with Disabilities' and the 'Act on the Honorable Treatment and Support of Veterans', individuals with disabilities and veterans are given preference during the recruitment process.Video interviews will include live coding tests and a brief assessment of ML fundamentals. Please prepare your PC environment for a smooth testing experience.

Who We AreJoin us in setting global standards for video understanding AI! Twelve Labs is dedicated to developing cutting-edge AI models specifically for video content, enabling efficient processing of vast amounts of video data. Our technology offers advanced capabilities for search, analysis, summarization, and generating insights from video.Our models are utilized by the largest sports leagues worldwide, quickly and accurately selecting highlights from extensive game footage, providing a hyper-personalized viewing experience. In South Korea, integrated control centers partner with us to efficiently analyze CCTV footage for rapid crisis response. Major broadcasters and studios across the globe leverage our models to create content for billions of viewers.Headquartered in San Francisco with an office in Seoul, Twelve Labs is a Deep Tech startup recognized for four consecutive years as one of the Top 100 AI Startups by CB Insights. We have secured over $110 million in funding from leading venture capital firms and corporations, including NVIDIA, NEA, Index Ventures, Databricks, and Snowflake. Our AI models are uniquely available through Amazon Bedrock, and we thrive on innovation and collaboration with exceptional colleagues worldwide.At Twelve Labs, we operate on core values that include:Honesty and reflection about ourselves and our teamsResilience and humility, embracing failure and feedbackA commitment to continuous learning and enhancing team capabilitiesIf you enjoy tackling challenging problems and growing through the journey, the opportunity awaits you here at Twelve Labs.About the TeamOur ML Data team operates on the belief that data determines AI model performance. We build high-quality data for training and evaluating multimodal AI models end-to-end. This includes gathering, filtering, processing, and labeling various types of multimodal data such as video, images, and audio. We collaborate with diverse teams to design datasets that unlock new model capabilities and develop evaluation datasets that reflect real user experiences. We also develop and continually enhance internal tools to perform these processes efficiently.The ML Data team plays a pivotal role in the development of Twelve Labs' world-class video understanding models through a meticulously designed data pipeline.About the RoleAs a Software Engineer specializing in Data, you will design and develop pipelines for multimodal (video, image, audio) data that fundamentally enhance model performance through data quality. If you have experience designing and operating distributed systems for handling unstructured multimodal datasets, you can make a significant impact in this position. The rigorously refined and accurately labeled data forms the foundation of all model development at Twelve Labs, and you will have the opportunity to influence model quality more than any other engineering role.We are looking for someone to help us build data infrastructure that elevates our video understanding technology to the next level.In this Role, You WillBuild data engines capable of collecting, preprocessing, refining, filtering, and labeling large multimodal (video, image, audio) datasets for LLM/VLM training.Design and develop data systems that efficiently manage and visualize petabyte-scale video, image, and audio data.Create libraries and services that deliver tangible impact beyond just eye-catching features.Collaborate closely with various teams to define project priorities and goals, leading technical initiatives from planning through development and operations.

Join Our Team!The Machine Learning Backend Engineer will be part of the ML Service Team, developing backend servers and common service platforms for various AI products within Toss Bank.The ML Service Team consists of Data Scientists, responsible for model development and application, and ML Engineers, who focus on server development, collaboratively creating innovative AI solutions.We work closely with product organizations responsible for banking services, leveraging AI technology to develop products. Collaboration with POs, designers, and developers ensures we manage the complete development cycle of AI products.Your Responsibilities:Design and develop robust server architecture to ensure a stable delivery of diverse AI products handled by the ML Service Team.Create individual services capable of handling large-scale traffic based on a solid server architecture.Develop servers that can efficiently serve heavy computational AI models with reliable response times and throughput.Operate services that allow distributed tracing in a microservices architecture while minimizing disruptions.We Are Looking For:Experience in server development and operation using one or more programming languages such as Python, Kotlin, or Go.Ability to design resilient service architectures and implement stable code structures.Experience deploying and monitoring servers in Kubernetes environments for stable operations is a plus.Experience and skills in developing and operating production services that handle high traffic or have stringent SLAs are desirable.A background in solving large and complex problems through software technology or creating business impact through services or products is advantageous.No prior ML experience is required, but a strong interest in AI/ML and a desire to gain experience in this area is welcomed.Resume Recommendations:Provide concrete examples of practical server development experience in your resume.Include details about technical considerations and your actual implementations while solving problems.If you have experience comparing costs and benefits among choices when balancing architectural complexity and development resources, please include that.Detail any projects that have made a business impact.Discuss experiences where you applied developed products in production and refined them based on performance metrics.Your Journey to Joining Toss Bank:Application submission > Live coding test > Job interview...

# About the Team- As an MLE (Commerce Recommendation) in the Toss Commerce domain, you will play a pivotal role in optimizing product visibility.- Our focus is on designing and enhancing recommendation models based on diverse data sources to present users with more relevant and appealing products.- This team leads the full spectrum of machine learning development, from problem definition to training, performance analysis, and enhancement.- We are building a product recommendation platform that provides meaningful shopping experiences for users by leveraging various ML techniques.- **Interested in learning more about Toss's Data Organization?** [→ *Toss Data Division Wiki*](https://recruit-data-division.oopy.io/)# Responsibilities- Develop models to predict product click-through rates (CTR), conversion rates (CVR), and other key metrics based on user behavior, product information, and contextual data.- Design and refine recommendation algorithms to optimize product exposure using predictive outcomes.- Conduct iterative experiments including model performance analysis, feature engineering, and hyperparameter tuning to enhance recommendation quality.- Quantitatively validate the impact of model improvements through various experiments and offline/online performance metrics.- Collaborate with domain experts and data analysts when necessary to accurately define and solve recommendation challenges.# Ideal Candidate- We prefer candidates with experience in developing recommendation systems or predictive ranking models.- Candidates who have designed and improved machine learning models for predicting user responses, like CTR and CVR, will be highly regarded.- Experience in experimenting, tuning, and analyzing models using various features is a plus.- Familiarity with major ML frameworks such as PyTorch, TensorFlow, and LightGBM is advantageous.- Candidates who have engaged deeply in problem definition and performance analysis beyond merely training models are welcome.- Strong communication skills and the ability to think data-driven while articulating complex problems clearly are essential.# Resume Tips- Detail any impactful projects you have undertaken that influenced your organization significantly.- Explicitly describe the problems you defined, the approaches you selected, and the methods you used in modeling-centered projects.- Highlight any quantitative problem-solving processes you have engaged in through iterative experimentation and performance analysis.# Journey to Join TossApplication > 1st Technical Interview (Coding) > 2nd Technical Interview > Cultural Fit Interview > Reference Check > Offer Negotiation > Final Acceptance- The first technical interview will include a simple coding test, resume review, and ML fundamentals assessment.- The second technical interview will focus on in-depth technical discussions and ML system design.# A Note for Potential Colleagues> "This role goes beyond mere modeling; it makes an impact in business."- The most fulfilling aspect of working at Toss is that we do much more than just modeling.- Previously, my tasks were limited to inputting data into existing models for performance evaluation, but now I am engaged in integrating unaggregated data into models and contributing to impactful solutions based on user understanding for our super app!

# Join Our Team- The AI Platform team is on a mission to create a platform that enables everyone to use AI technology quickly and reliably. We technically support the use of AI across Toss.- We are developing the necessary tools and platforms for new AI systems such as Retrieval-Augmented Generation, Agents, and Assistants to be rapidly experimented with and reliably operated.- The platform we create is not just a simple toolkit; it is designed with scalability in mind, allowing AI technology to be implemented by more teams effectively.- Addressing unresolved problems means that collaboratively defining and structuring technical directions is crucial.- **Want to learn more about Toss’s data organization?** [→ *Toss Data Division Wiki*](https://recruit-data-division.oopy.io/)# Responsibilities- Integrate LLM-based components such as Retrieval, Generation, and Vector Search into a platform that various teams can reuse.- Provide features that integrate both SaaS and self-hosted LLMs and ensure stable operations.- Design the foundation for creating and experimenting with Agent systems more easily, including prompts, tools, and context configurations.- After experiments, ensure stable operation in the service by systematizing and tooling the serving and operational flow of RAG and Agents.- Develop a foundation to quantitatively evaluate the performance and quality of Agents and provide it as a platform.- Design a common environment and experience to enable rapid experimentation and application of AI systems not only within the team but also across other teams.- Structure unformatted technical elements and create directions that can expand into broader problems.# We Are Looking For- Individuals with experience applying technologies such as LLM, RAG, and Agents to real-world problems.- Those who can technically define unstructured problems and systematically solve them.- Candidates who have collaborated with multiple teams to develop and operate technology as if it were a product.- Those who can quickly adapt to new technological trends and integrate them naturally within the team.- Individuals interested in simplifying complex AI systems into a consistent and straightforward user experience.# Preferred Qualifications- Experience designing RAG components such as Retrieval, Generation, and Vector Search independently and integrating them at the system level.- Experience selecting and operating various LLM serving structures (OpenAI API, HuggingFace, vLLM) tailored to service situations.- Experience structuring various purpose-driven Agents and applying them in service operations is highly welcomed.- Experience proactively designing experimental environments or tools based on the requirements of platform users (internal developers, model engineers, etc.).- Experience creating and operating common platforms or RAG-based systems that can be extended across multiple projects or domains is especially desirable.# Resume Tips- If your past projects had significant impacts on the organization, please detail them.- Rather than just listing languages, platforms, frameworks, or technologies used, provide context about the project's objectives, the methods you employed, and how you solved the problems.- Include experiences where you resolved critical issues during platform operations or optimized performance/resources.- If you have contributed to open source by fixing bugs or enhancing functionalities, please share those experiences.# The Journey to Join Toss- Application submission > Job interview > Cultural fit interview > Reference check > Compensation negotiation > Final acceptance- The job interview will feature in-depth technical interviews focused on ML system design.# A Message for Future Colleagues“We don’t just serve rapidly evolving AI models; we build systems that ensure these models operate stably and are continuously improved.”- The AI Platform team is responsible for the serving, experimenting, and operating infrastructure of various AI technologies such as LLM-based services, RAG systems, and search infrastructure to ensure smooth operation in production environments.- We efficiently manage GPU resources and clusters, utilizing vLLM, Triton, Model Registry, etc., to automate experiments and deployments.

Who We AreWe are on the lookout for exceptional talent to help shape the global standard of video understanding AI!At TwelvaLabs, we are creating world-leading AI models specifically designed to process vast video datasets, enabling specialized functionalities such as search, analysis, summarization, and insight generation.Our models are utilized by the largest sports leagues globally to swiftly and accurately curate highlights from extensive game footage, providing an ultra-personalized viewing experience. In South Korea, integrated command centers rely on TwelvaLabs to efficiently navigate CCTV footage for rapid crisis response, while major broadcasters and studios worldwide leverage our models to produce content for billions of viewers.TwelvaLabs, a deep tech startup with offices in San Francisco and Seoul, has been recognized for four consecutive years as one of the top 100 AI startups globally by CB Insights. We have secured over $110 million in investments from premier VCs and companies including NVIDIA, NEA, Index Ventures, Databricks, and Snowflake. Our AI models, uniquely developed in Korea, are exclusively serviced through Amazon Bedrock. We are committed to creating innovative products alongside outstanding colleagues and growing together with our global clientele.At TwelvaLabs, we operate based on core values that include:A commitment to honesty and self-reflection about ourselves and our teamResilience and humility in the face of failure and feedbackA dedication to continuous learning to enhance the team's capabilitiesIf you enjoy solving challenging problems and growing through the process, the opportunity awaits you here at TwelvaLabs.About the TeamYou will be joining the team responsible for multimodal representation learning and production serving at TwelvaLabs. We develop models that integrate various modalities such as video, audio, and text into a unified embedding space, reliably serving them in production systems used by thousands of clients worldwide.Our responsibilities include conducting experiments on multimodal embedding models within a large-scale distributed learning environment and transitioning research outcomes into real-time inference systems. With access to top-tier GPU resources like the NVIDIA B300, we minimize the transition cycles from research to production.Our research results reach global clients within months, and we collaborate closely with the Research, Product, and Infrastructure teams to create significant technical impacts.About the RoleAs a Senior Machine Learning Engineer on the Embedding & Search team, you will play a pivotal role in developing and owning critical components of TwelvaLabs' search and retrieval platform. This platform integrates vector search, lexical retrieval, and reranking to deliver fast, accurate, and scalable search experiences for our clients.This position demands expertise in systems-heavy ML engineering, focusing on information retrieval, ML serving, and distributed systems. We seek a strong engineer capable of deconstructing well-defined problems with moderate ambiguity into concrete milestones, delivering reliable and high-performance solutions.Key ResponsibilitiesDevelop and manage essential subsystems of our search platform on EKS, including vector indexing (ANN), lexical retrieval, hybrid fusion, reranking, and temporal (segment-level) search.Enhance retrieval performance across vector and lexical paths at a scale ranging from millions to billions.

Join Our Innovative TeamThe Machine Learning Engineer (Infra) will be part of the ML Platform Team within the Product Division at Toss Securities.The primary goal of the ML Platform Team is to create an optimal machine learning platform that enables the efficient and stable development and operation of various AI/ML services at Toss Securities.The ML Engineer (Infra) will focus on maximizing the efficiency of large-scale AI infrastructure, finely controlling resource usage, and enhancing infrastructure performance to its peak. Your Responsibilities Design and operate high-performance AI computing environments reliably.Design and operate top-of-the-line GPU clusters (H100, B300 series) connected via InfiniBand and high-performance storage (400Gbps) within a Kubernetes environment.Beyond merely building infrastructure, optimize networks and storage to extract the full potential of hardware performance. Develop a comprehensive control system for the entire AI infrastructure.Create an observability system to integrate and monitor AI resources distributed across internal infrastructure and external cloud.Implement management features to prevent resource monopolization by specific services and allocate resources precisely based on importance. Create automation tools for the most efficient resource usage.Analyze actual usage patterns to develop tools that recommend 'just-right resources' to avoid waste.Implement features that automatically scale up or down based on real-time model performance or error rates, and reallocate GPUs where necessary. Establish an environment for identifying and resolving model performance bottlenecks.Build profiling environments to accurately pinpoint slowdowns during model training or serving.Support the analysis and improvement of performance degradation causes between hardware and software. Who We Are Looking ForYou have experience building and operating Kubernetes-based ML infrastructures that handle large-scale traffic.You take responsibility for reliably operating live services beyond simple development.You have experience persistently analyzing and debugging to resolve root causes when issues arise.You possess a solid understanding of system resources (GPU/CPU/Memory/Network/Storage) and have experience building monitoring systems for them.You value the process of solving various problems that arise during service operations and strengthening the system. Preferred QualificationsExperience in unified monitoring of resource usage in large-scale clusters.Experience building systems to systematically control resources through Quota and Rate Limits.Experience with open-source platforms like Kubeflow or Kubernetes, including in-depth modifications as needed.Experience analyzing and optimizing bottlenecks at the kernel level using tools like Nsight Systems/Compute or PyTorch Profiler.Experience designing tasks to reduce costs or enhance performance tailored to workload characteristics (Rightsizing, Cost Optimization).Experience leveraging GPU virtualization technologies like MIG and MPS to maximize resource utilization.

Who We AreAt TwelveLabs, we are at the forefront of creating innovative multimodal foundation models capable of interpreting videos in a manner akin to human understanding. Our groundbreaking models have set new benchmarks in video-language modeling, granting us enhanced capabilities and fundamentally reshaping how we engage with and analyze diverse media formats.With over $110 million in Seed and Series A funding, we are supported by leading venture capital firms including NVIDIA’s NVentures, NEA, Radical Ventures, and Index Ventures, alongside esteemed AI pioneers like Fei-Fei Li, Silvio Savarese, and Alexandr Wang. Our headquarters are in San Francisco, complemented by a significant presence in the APAC region from our Seoul office, reflecting our dedication to fostering global innovation.Our strategic partnerships with NVIDIA and AWS provide us with access to state-of-the-art hardware, including B300s, enabling us to explore the frontiers of video AI technology.As a global organization, we celebrate the distinct journeys of each individual. Our diverse cultural, educational, and life backgrounds empower us to consistently challenge traditional norms. We seek passionate individuals who resonate with our mission and are eager to contribute to transformative advancements in technology. Join us in revolutionizing video comprehension and multimodal AI.About the TeamThe Pegasus team is integral to TwelveLabs' video understanding initiatives and spearheads the development of Pegasus, our Video Analysis product. We focus on constructing multimodal video analysis systems that excel in instruction following and generate complex, hierarchically organized outputs. Our priority is to deliver products that hold real-world significance rather than engaging in isolated research, collaborating within a goal-driven, cross-functional team comprising both ML researchers and engineers.Our responsibilities encompass a wide array of challenges: large-scale distributed training of multimodal LLMs from pre-training to reinforcement learning, precise temporal segmentation and structured metadata extraction for practical applications, extending temporal context lengths to several hours, and implementing data curation processes that facilitate well-aligned evaluations and performance enhancements through improved training data.Our team utilizes the latest cutting-edge chips, including NVIDIA B300s, to propel the limits of video analysis systems—accelerating our transition from research to production as swiftly as possible.

Who We AreAt Twelve Labs, we are at the forefront of developing state-of-the-art multimodal foundation models that enable video comprehension akin to human understanding. Our innovative models have set new benchmarks in video-language modeling, enhancing our capabilities and revolutionizing how we engage with and analyze diverse forms of media.Securing over $110 million in Seed and Series A funding, we are supported by prestigious venture capital firms including NVIDIA’s NVentures, NEA, Radical Ventures, and Index Ventures, along with leading AI pioneers like Fei-Fei Li, Silvio Savarese, and Alexandr Wang. With our headquarters in San Francisco and a strong presence in Seoul, we are committed to fostering global innovation.Our collaboration with NVIDIA and AWS equips us with cutting-edge chips, including B300s, allowing us to expand the horizons of video AI technology.We embrace the uniqueness of each individual's journey, believing that our diverse cultural, educational, and life experiences enhance our ability to challenge conventional thinking. We seek motivated individuals who are passionate about our mission and eager to make a significant impact as we advance technology to transform the world. Join us in revolutionizing video understanding and multimodal AI.About the TeamThe Pegasus team is pivotal to Twelve Labs' video understanding services, spearheading the development of Pegasus, our Video Analysis product. We focus on creating multimodal video analysis systems that excel in instruction-following capabilities and generate complex, hierarchically structured outputs. Our emphasis is on delivering products with tangible real-world value, working within a goal-oriented, cross-functional team comprising both machine learning researchers and engineers.Our work addresses a wide array of challenges, including large-scale distributed training of multimodal LLMs from pre-training to reinforcement learning, precise temporal segmentation, and structured metadata extraction for practical applications. We also enhance temporal context length and refine data curation processes to align evaluation and performance improvements through enhanced training data.Our team has access to the latest advanced chips, such as NVIDIA B300s, which accelerates our transition from research to production at an unprecedented pace.

# Join Our TeamThe Machine Learning Engineer (Service) is part of the ML Tribe within Tossbank's Data Division.Our team is dedicated to leveraging AI (LLM, ML) technologies to tackle challenges previously unresolved in the financial sector, optimizing operations with the aim of maximizing profitability.Your ResponsibilitiesDevelop and operate agents to optimize financial tasks.Design and develop systems for delivering robust and scalable financial ML models.We Are Looking For Someone WhoHas experience and skills in developing and operating ML services with high SLA requirements in high-traffic environments.Can design service architectures that consider both stability and scalability, alongside writing maintainable, high-quality code.Has experience or interest in optimizing tasks with agents.Is familiar with technologies such as FastAPI, LLM, Agents, Kafka, Spark, Flink, and Airflow.Has a track record of utilizing AI technologies to solve large and complex problems or has created services/products that have made a significant business impact.Resume TipsStructure your resume around practical development experiences with specific examples.Detail your technical considerations and implementations aimed at solving problems.Discuss your experiences in analyzing trade-offs between architectural complexity and development resource allocation to determine cost efficiency.If you have worked on projects that resulted in business impacts, please elaborate on those experiences.Share your experiences in applying developed products to production and refining them based on performance metrics.Journey to Joining TossbankApplication Submission > Live Coding Test > Technical Interview > Cultural Fit Interview > Reference Check > Offer Negotiation > Final Acceptance and OnboardingImportant NotesJob offers may be rescinded if any false information is found in your resume or if disciplinary actions are noted in your employment history.According to Tossbank's Employment Regulations Article 8 (Hiring Disqualifications), candidates found to have disqualifying reasons may have job offers revoked.Applicants with disabilities or those eligible for national veterans' benefits are given preferential treatment in accordance with relevant laws.

About UsAt TwelveLabs, we are at the forefront of innovation, developing advanced multimodal foundation models capable of understanding videos as humans do. Our groundbreaking work in video-language modeling has set new benchmarks, enhancing our ability to interact with and analyze diverse media forms.With over $110 million secured in Seed and Series A funding, we are supported by prestigious venture capitalists, including NVIDIA’s NVentures, NEA, Radical Ventures, and Index Ventures, alongside notable AI leaders like Fei-Fei Li and Silvio Savarese. Headquartered in San Francisco, our significant presence in the APAC region, particularly in Seoul, illustrates our dedication to global innovation.Leveraging partnerships with NVIDIA and AWS grants us access to state-of-the-art technology, enabling us to explore the full potential of video AI.We celebrate the uniqueness of every individual's journey, believing that our diverse backgrounds foster an environment where we can challenge norms and push technological boundaries. We invite driven individuals who are passionate about our mission to join us as we transform the landscape of video understanding and multimodal AI.The TeamThe Pegasus team is central to TwelveLabs' video comprehension capabilities, focusing on our Video Analysis product. We are dedicated to developing multimodal video analysis systems that excel in high instruction-following capabilities and produce complex, structured outputs. Our emphasis is on delivering products that create real-world impact, collaborating in a cross-functional team of ML researchers and engineers.Our challenges include large-scale distributed training of multimodal LLMs from pre-training to reinforcement learning, precise temporal segmentation, and structured metadata extraction for practical applications. Additionally, we are involved in extending temporal context lengths and enhancing training data for improved evaluation and performance.Equipped with the latest NVIDIA B300 chips, our team accelerates the transition from research to production, striving for rapid advancements in video analysis systems.

Join our innovative team at Tosscareers as a Machine Learning Engineer specializing in Natural Language Processing (NLP). In this role, you will harness the power of machine learning to develop state-of-the-art NLP models that enhance user experience and drive our product forward. Collaborate with cross-functional teams to integrate advanced algorithms and contribute to groundbreaking projects that reshape how our users interact with technology.

Join Our Team! As a Machine Learning Engineer specializing in LLM, you will be a vital part of the TossBank ML Platform Team. Our ML Platform Team is dedicated to developing and operating the machine learning platform for TossBank. Your Responsibilities: Utilize LLM serving frameworks to directly operate and optimize models in a GPU environment. Build and manage the LLMOps platform, including LLM gateway servers, workflows, and vector databases. Develop LLM applications using various LLM technologies. Who We Are Looking For: Experience in serving models using frameworks like vLLM, SGLang, and TensorRT LLM in high-traffic environments is a plus. Previous experience in building and managing LLMOps platforms is preferred. Hands-on experience in developing LLM applications using LLM technologies is desirable. Familiarity with LLM fine-tuning and optimization is advantageous. A passion for the latest LLM technologies and a willingness to embrace new challenges are essential. Resume Tips: Detail impactful projects from your past experiences. Expand on your experience with LLMOps platform construction and troubleshooting. If you've implemented improvements in real services, share quantifiable results (omit sensitive information as necessary). Path to Joining TossBank: Application submission > Live coding test > Job interview > Cultural fit interview > Reference check > Compensation negotiation > Final acceptance and onboarding. Important Notes: Any false information in resumes or disciplinary records from previous employment may result in cancellation of recruitment. Applicants who fall under disqualifying criteria as per TossBank's hiring regulations may also face cancellation of recruitment. Priority is given to applicants with disabilities and those eligible for national veterans' benefits in accordance with applicable laws. A Word from a Colleague: "This company is the best for ML Engineers who seek explosive growth." It provides a fantastic environment to work freely and immerse in tasks. It's a great choice for those looking to grow alongside exceptional colleagues.

About the Team You'll JoinThe Machine Learning Engineer (Product) is an integral part of the team responsible for AI services at Tossbank, collaborating with a diverse group of professionals including ML Engineers, Data Scientists, Server Developers, and Product Owners to oversee AI product development from inception to deployment.You will collaborate with various squads utilizing AI services to create innovative products and solutions.Your primary objective will be to automate and enhance processes such as identity verification, document processing (OCR), and customer service automation using cutting-edge AI technology.Your ResponsibilitiesDesign and develop serving systems to ensure the stability and scalability of various machine learning models, including advanced AI technologies like Vision and LLM.Architect optimal serving structures tailored to specific ML model needs, implementing rapid and efficient solutions using specialized tech stacks.Create individual services capable of handling high traffic volumes based on a robust server architecture.Design serving pipelines for both batch and real-time ML processes, ensuring reliable operations with large datasets.Develop servers that can reliably serve heavy computation AI models while maintaining optimal response times and throughput.Preferred QualificationsProven experience in deploying ML models in real-world environments, focusing on performance tuning and optimization.Experience in developing and operating large-scale ML services that meet high SLA requirements within production settings.Familiarity with model compilation and optimization tools like TensorRT and ONNX Runtime to maximize inference performance on GPU environments.In-depth understanding and experience with ML serving and server development technologies, including Triton Inference Server, Kubernetes, and FastAPI.Strong knowledge of large-scale distributed systems and data pipeline technologies such as Airflow, Hadoop, Kafka, Spark, and Flink.Ability to design service architectures that balance stability and scalability while producing maintainable, high-quality code.Experience in leveraging AI technologies to solve complex problems through innovative products or services that drive business impact.Resume RecommendationsHighlight practical development experience with specific, case-driven examples.Include technical insights into problem-solving processes along with actual implementations.Discuss experiences prioritizing cost-effectiveness among trade-offs related to architectural complexity and development resources.